Stopping a cyberattack while walking your dog - defensive AI security CEO says it's not ruff to do
Corma CEO tells The Reg it's building 'One ring to rule them all, for the defenders to have this power'
Alon Pluda, CEO of AI security startup Corma, says his company aims to address the defensive gap in cybersecurity. This gap refers to models excelling at offensive security but struggling with defensive tasks. Pluda shared a story of a customer who received a notification from a Corma agent while walking his dog, alerting him to a live attack.
The agent blocked the malware, mitigating the intrusion in under 10 minutes. Pluda's startup received $60 million in funding and is working with Fortune 100 companies to train models for superintelligence in defensive cybersecurity. The company tested four frontier models against each other, finding that they could successfully implant a backdoor in 85 percent of scenarios but only detected 19 percent of attacks.
Pluda attributes this imbalance to the data models are trained on, which lends itself to offensive security more than defensive tasks. Corma's AI agents, powered by their models, operate across various defensive security tasks, reducing threat response times by over 94 percent and expanding security coverage by 15 times.
Written by urgent.news from The Register's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
